Kirk's Hammer: Bayesian Thinking

Alright, so I don’t know if anyone reading this has heard, but there’s been a bit of a poker scandal on gambling Twitter over the last week. I know everyone is dying for more analysis and conversation around the scandal; sadly, that is not what I am going to be talking about here.

But I clearly have some form of mental illness that compels me to argue with people I have zero chance of convincing of my point of view. And in many of these conversations, I have been called a flip-flopper because I’ve changed my opinion. These fine folks that call me that are typically saying it in an insulting tone, but honestly, I take it as a big compliment.

Changing your mind is awesome. The first time I was on Circles Off, my +EV move was to change my mind more, and I believe that at the time I didn’t change my mind nearly enough—and I still don’t change my mind nearly enough.

Someone said to me recently, “If you don’t look back at yourself a year ago and think they’re an idiot, you’ve done something wrong.” And I think that’s entirely true.

What Is Bayesian Thinking?

Bayesian thinking is a way of updating what you believe when you get new information.

1. Start with a Prior

This is what you already believe about something—your best guess or probability before seeing new evidence.

Example: You have a co-worker you’ve worked with for 3 years; you guys are work friends but don’t know each other outside the office. You are trying to determine whether he is trustworthy. He has never explicitly broken trust with you or anyone around you, but also, you haven’t ever really confided in him.

OK, so let’s set a prior: you don’t know this person well, but you’re friendly. You don’t really know anyone he knows because you only work together. You have nothing to determine him to be untrustworthy, but nothing solid to point to him being clearly trustworthy. You’re generally trusting, so you think there’s a roughly 85% chance this person is trustworthy.

2. Get New Evidence

Now you get a new piece of info: you told him something that you thought was private, and he told another co-worker.

3. Update Your Belief (Posterior)

You now have a new piece of info, and it’s probably the most pressing piece of info you have. But you didn’t explicitly tell him it was a secret, and it wasn’t a huge deal—but it was pretty obviously something you wouldn’t want shared.

I would move my prior pretty significantly here, maybe to a 40–50% chance, and I would probably avoid telling him info I didn’t want to get out.

The Value of Mental Models

A criticism here is that these percentages aren’t perfect—that’s obvious. But having a mental model for most things in life puts you way ahead of anyone who doesn’t.

Most people process that same example as a 100% chance the person is trustworthy, and then a 0% chance after the breaking of trust. Using Bayesian thinking blows that kind of binary reasoning out of the water.

People also might think that it takes a lot of time to think that way, but if you train yourself to not think in certainties, you will learn to very quickly be able to shift priors almost subconsciously.

Win Probability as a Metaphor

I think another great way to think about it is using this random win probability chart that ChatGPT cooked up.

No sane person would ever call you a flip-flopper for thinking Team A was unlikely to win at minute 4, then likely to win at minute 9, then very unlikely to win at minute 15, then almost certain they’d win at minute 20.

Life is no different—it just happens slower. You get information and change your opinion. Sometimes those swings can be 1–2%, and sometimes they can be 40–50%, but very rarely should they be 0% to 100%.

Certainty Is a Trap

The last thing I’ll mention is that people will often say something like, “If I were in situation X, I would have never done Y,” as an explanation for why someone is lying. This is horrible reasoning.

There are countless things every day I witness other people do that I would not do—but that’s because humans are different and react differently to similar situations.

A much better framework would be: “I think most people likely wouldn’t act that way. It’s not impossible that someone could act that way, but I deem it unlikely.” Allowing for no nuance, and speaking in certainties, often opens you up to looking like a fool.

--

Sign up to Underdog and deposit using code CIRCLES to receive up to $1K in bonus cash + a gimme pick or simply click this link: http://play.underdogfantasy.com/circles

Get yourself an Edge Boost Visa Debit to eliminate the headaches of deposits and for cash back on your betting: https://www.edgeboost.bet/guide/?oid=196&affid=117